- Scaleable to billions or trillions of nodes
- Runtime agnaustic orchestrator, will support OCI initially.
- Control-plane vs worker node == no difference for install/user.
- Single binary install
- Based on promise theory?
- HA on as few as 3 nodes
- Gitops core?
- Service mesh
- Other things
- Initial Cluster API
- Node to node cluster communication interface
- Node to node cluster communication implementation (zeroMQ)
- Container Runtime integration
- Resource Defs
- container?
- kube pod?
- kube deployment?
- kube service?
- CP Storage
- CP Consensus
- Machine taints
- Network
Yes, not only will I be streaming it a few nights a week on twitch starting later this year. I am going to make youtube videos covering the "how and whys" of the project. Because of this there are a few constraints this puts on the branching of the project. That being said, I would stay away from this project if I was you until at least 2023. And then watch the youtube series first and then join the live streams and THEN start contributing :)
There are 3 types of branches
- Main - this is the main branch for development
- vX.X - this is a release branch for the release of x.x
- ytX - this is a branch that contains spesific code for youtube episode x
Building and local unit testing are made simple by cargo.
Requires both cmake and protocal buffers installed. Want to help out with this project? add more instructions here :)
cargo build
use the --release flag for a more optimized build. It will take longer to compile.
cargo check && cargo fmt -- --check && cargo clippy
It could be required for you to use rustup component add clippy && rustup component add fmt to use those commands.
Currently there are a few failures in check and clippy; in the future both of these will be required to pass to all of these to be merged.
- ClusterAPI app
- Node Communication Interface
- Node Communication implementation
- Container Runtime integration
Based on Promise Theory. designed to be control-plane-less. All nodes created equal. info
Play nice with others please. info
Here is a best effort deployment guide.